Artificial intelligence has become one of the hottest areas of tech in recent years. And one of the subsectors of AI that has received the most attention is called machine learning. Machine learning uses systems to perform tasks without explicit instructions. It builds a mathematical model of sample data, using that to make predictions or decisions. With machine learning’s increasing importance, investors have increasingly focused on machine-learning stocks.

Machine learning requires both hardware and software to function efficiently. Hence, several different types of tech companies have both researched and deployed machine-learning capabilities. It already plays a critical role in applications such as internet search, data analytics, social media and home assistants. As self-driving cars and 5G reach the marketplace, the role of machine learning will only increase.

The following machine-learning stocks have played critical roles in this technology and could become some of the best stocks to invest in for this growing trend:

Alphabet (GOOGL, GOOG)

Perhaps no company defines machine-learning stocks better than Google parent Alphabet (NASDAQ:GOOGL, NASDAQ:GOOG). Though Alphabet involves itself in many areas, it has become one of the more prominent machine-learning stocks in recent years. In 2015, Google introduced RankBrain. RankBrain transformed a search engine that had previously operated under a strict set of rules. Now, with RankBrain, Google search employs neural nets to analyze data to anticipate and learn behaviors comparable to a human brain.

Considering its reach and its growth rate, GOOGL stock trades at reasonable levels. It trades at a forward P/E ratio of 23.2. Moreover, it appears on track to continue the double-digit growth rates it has enjoyed for years to come. Analysts expect 2018 profit growth to come in at 30.6%. While that will come down in 2019, Wall Street predicts 12.4% growth for this year. It has increased by 10% since hitting its 52-week low on Christmas Eve. This move took GOOGL stock out of bear market territory, and it currently trades at about 16% below its 52-week high.

The biggest glaring fault I see with GOOGL stock involves the lack of a dividend. Alphabet maintains a $751 billion market cap, and it held over $106 billion in cash as of the end of the third quarter. Given that situation, I think the pressure will mount on the company to offer a payout. If they finally begin to provide a payout, I believe the stock would rise further.

Whatever decision Alphabet’s board makes, Google will maintain its search dominance for the foreseeable future. For those who want a large-cap stock defined largely by machine learning, no mega-cap equity comes close to Alphabet.

Apple (AAPL)

Admittedly, Apple (NASDAQ:AAPL) might seem like a strange choice even if it is one of the more important machine-learning stocks. Wall Street has hammered AAPL stock in recent months as disappointing iPhone sales sent revenue and profit estimates plunging.

Investors may have to adjust to a long-term slowdown in iPhone sales. Nonetheless, iOS will remain one of the principal ecosystems in the smartphone market today. To maintain that position, it has worked to integrate machine learning into its ecosystem. Examples of this include Siri, Camera and QuickType.

Due to the iPhone driven decline in AAPL stock, it has reached bargain valuations. Thanks to a 33% decline from the 52-week high, AAPL stock now trades at a forward PE ratio of 11.7. AAPL has also seen a history of bouncing back from such lows.

Downward revisions have weighed on profit growth. However, analysts still expect the company to return to double-digit profit growth for next year. Current growth estimates for fiscal 2020 stand at 11.5%. Investors will likely bid AAPL stock higher again once that growth returns. While Apple has become a hated name in tech, iOS still enjoys wide use. Hence, we can expect AAPL stock to recover, and machine learning will help its popularity return.

Facebook (FB)

Like Apple, Facebook (NASDAQ:FB) has become one of the less popular machine-learning stocks in recent months for unrelated reasons. Data privacy concerns and slowing revenue growth weighed on the company since last summer.

Despite the pessimism surrounding the company, it has conducted extensive research in machine learning. It uses the technology to rank feeds, ads, and search results. It also uses machine learning to prevent spam. Machine learning also plays a critical role in translating languages and “reading” images, allowing blind people to utilize the social media site.

It also happens that FB stock trades at 2017 levels. The recent slump began with the largest one-day market cap loss in stock market history. The stock would go on to lose about 44% of its market value. However, the stock may now have begun to recover. Since Christmas Eve, the stock has risen by over 15%. Still, despite the move higher, FB stock trades at about 20 times forward earnings. Although earnings will stagnate in the current year, analysts forecast a return to double-digit earnings increases next year. Over the next five years, they predict average annual earnings growth of 17.7%.

FB stock may struggle for now as it faces legal scrutiny over data privacy concerns. However, the disruption appears temporary. Soon, double-digit profit growth will again return. Machine learning will likely power much of that increase.

Intel (INTC)

Like many tech giants, Intel (NASDAQ:INTC) has also embraced machine learning. The one-time leader in PC hardware struggled to redefine itself following the move to smartphones and tablets. However, with a new specialty in data centers and IoT, machine learning has become a critical component in the new Intel. Intel Select Solutions helps to speed up the deployment of data center solutions, enhance data mining and analytics, and optimize database performance.

One solution is what it calls Optane DC Persistent Memory. Optane allows it to store more of what it calls “hot” data closer to the CPU. This means that the systems can deliver results faster and improve efficiency.

Despite this new focus, Wall Street has only taken notice slowly. At least from a valuation perspective, INTC stock has become the most inexpensive of the major machine-learning stocks. Its forward P/E now stands at around 9.9. It did not help that the stock fell by about 5.5% after missing estimates on revenue.

For now, one will get what they pay for as the current glut in chips will lead to negative profit growth for 2019. However, in fiscal 2020, analysts expect to see 4.6% profit growth. They also predict that that growth rate will return to double-digit rates after 2020. Still, once chip supplies come into balance with demand, INTC stock should find itself well positioned to benefit from machine learning and rise to higher multiples.

Nvidia (NVDA)

In recent years, Nvidia (NASDAQ:NVDA) has made a name for itself by parlaying its graphic and gaming capabilities into various AI applications. One of those AI applications, of course, is machine learning. It has developed what it calls the Rapids suite. This collection of software libraries allows for the execution of analytics entirely on Nvidia GPUs. This serves as a critical component to Nvidia as its chips power autonomous cars and other AI-related applications.

NVDA stock may have also become a buying opportunity for investors. Lowered earnings guidance in a recent report took NVDA down by more than 15% in a single day. As a result, NVDA trades at a forward P/E of just under 19, lows it has not seen since the middle of the decade.

Like other chip stocks, the decline of crypto and the sudden oversupply of chips has hurt profit growth temporarily. As a result, profits will fall in the upcoming fiscal year. However, once supply and demand find balance again, double-digit profit growth should resume. Analysts predict average profit growth of 15.1% per year over the next five years.

Still, even with its short-term struggles, NVDA stock has arguably become one of the best stocks in the chip industry. Once the chip glut is behind the company, NVDA will rise again, and it will continue to hold its position as one of the more critical machine-learning stocks.

As of this writing, Will Healy did not hold a position in any of the aforementioned stocks. You can follow Will on Twitter at @HealyWriting.